serofit {denvax}

Model fitting for serological data

Description

Model fitting for serological data

Usage

serofit(seropositive, N, age.min = use.default(1L:length(seropositive),
  "age.min"), age.max = use.default(age.min, "age.max"))

Arguments

seropositive

the number of seropositive samples for each age group; length(seropositive) must be at least 3

N

the total number of samples for each age group; length(N) must equal length(seropositive)

age.min

the low age in age groups; defaults to '1:length(seropositive)', i.e. assumes the seropositive data corresponds to yearly cohorts starting at age 1.

age.max

the upper age in age groups; defaults to 'age.min', i.e. assumes each category corresponds to a single year

Details

Fits a constant force of infection, two-risk category model using seroprevalence survey data. i.e.:

$$ P_+(A) = p_H * (1-(1-f_H)^A) + (1-p_H) * (1-(1-f_L)^A) $$

This probability is fit to the seroprevalence by age category data, using maximum likelihood and optim.

Value

a list of best-fit parameters, all numeric values:

f_H

force of infection, for the high risk group

f_L

force of infection, for the low risk group

p_H

the proportion of the population at high risk

Examples

require(denvax);
data(morrison2010) # has counts by age
fit <- with(morrison2010, serofit(sero=Seropositive, N=Number, age.min=Age))
if (requireNamespace("data.table", quietly = TRUE)) {
data(lazou2016) # has counts by age range, instead of counts for every year
# this example uses `data.table`` functions to simplify processing
# several groups at once
  lazou2016[,{
    agerange <- data.table::tstrsplit(Age, "-")
    serofit(
      sero     = Seropositive,
      N        = Number,
      age.min  = as.integer(agerange[[1]]),
      age.max  = as.integer(agerange[[2]])
    )
  }, by = Country]
}